DESCRIPTION: (Applicant's Abstract) AVMs represent a small proportion of the total incidence of stroke but typically affect otherwise healthy young adults, usually in the productive years of life. The primary reason to treat cerebral arteriovenous malformations (AVMs) is the prophylaxis against new or recurrent intracranial hemorrhage (ICH). However, treatment of AVMs by surgery, endovascualar embolization and radiosurgery carries a significant risk. Further, since ICH prevention is the main impetus to treat AVMs, treatment risk for complex lesions should be less than the projected cumulative risk for spontaneous ICH. In our preliminary work we have strong evidence that certain angiographic and physiologic factors are related to an initial presentation of ICH. Further, in a prospective preliminary study, we have shown that initial presentation of with ICH is related to recurrent hemorrhage, which supports the notion that factors associated with the presenting ICH should be predictive of recurrent ICH. The general objectives of this study are 1) to complete a prospective observational study of recurrent spontaneous AVM hemorrhage; 2) to explicitly define the clinical, morphologic, and hemodynamic criteria that are associated with the initial presentation of ICH; and 3) to identify a reliable set of risk factors to predict ICH. We hypothesize that there exists a set of distinct physiologic and anatomic conditions that increase the risk for spontaneous ICH. Further, we hypothesize that a number of factors are associated with a low incidence of ICH, including a) large AVM size, b) low intranidal pressure, c) lack of periventricular deep venous drainage and venous stenosis and d) lack of dysfunctional dilation, i.e., aneurysm formation. Because, paradoxically, such patients at low risk from the natural history of the disease may often by at the highest treatment (surgical) risk, this group needs to be identified for optimal outcome management. Therefore, we propose studies on 120 AVM patients over a three-year period. Using an established databank, we will collect data from clinical course, standard imaging (angiography, CT, MRI), and physiologic (intravascular pressure) characteristics. We will develop a statistical model, based on explicit anatomic and physiologic criteria, for prediction of spontaneous ICH that can be weighed against treatment risk. We will also be able to add the proposed 120 new patients to our existing prospective sample (n=150) to complete the observational study of new or recurrent spontaneous AVM hemorrhage. At the conclusion of the grant period, our goal is to be in a position to construct and justify rational clinical trials that will randomize patients with high treatment risk and what we will determine to be low natural history risk into conservative versus aggressive therapy groups, including some combination of non-interventional, endovascular, radiosurgical or surgical treatment. The project will lead to a better understanding of the pathophysiology and natural history of cerebral AVMs and a more cost-effective use of medical resources by allowing more rational selection of patients for complex and multi-modal treatment.